PA RT Y P O L I T I C S V O L 1 2 . N o . 2 pp.

211–232

Copyright © 2006 SAGE Publications London Thousand Oaks New Delhi

www.sagepublications.com

IMPLEMENTATION OF POLITICAL
PARTY GENDER QUOTAS
Evidence from the German Länder 1990–2000

Louise K. Davidson-Schmich

ABSTRACT

The use of gender quotas to increase women’s political representation

in legislatures has expanded in recent years. Scholars have identified
three main categories of gender quotas, including constitutional,
election law and political party quotas. While considerable attention has
been devoted to classifying types of quotas, to explaining why they were
adopted and to investigating which quota provisions are most effective,
little is known about the actual implementation of political party gender
quotas. This article is a first step in explaining when parties live up to
their promises to promote women’s representation and when they fail
to do so. Focusing on Germany, I examine four parties’ experiences in
implementing quotas in 16 states over a decade. I argue that while
structural factors such as the electoral and party systems as well as the
supply and demand of female candidates are not unimportant, cultural
variables best explain whether or not parties actually comply with their
own quotas.

KEY WORDS
gender quotas
party quotas
women in German politics
women
in legislatures
women’s representation

Introduction

Both the use of gender quotas to increase women’s political representation

in legislatures around the world and the academic study of this practice have
expanded in recent years. Scholars have identified three main categories of
gender quotas (Global Database, 2003). First, 14 countries have adopted
‘constitutional quotas’ in which the country’s constitution mandates a
particular kind of gender representation. For example, Uganda’s requires
one female representative from every district of the country. Second, 31
1354-0688[DOI: 10.1177/1354068806061338]
 PA RT Y P O L I T I C S 1 2 ( 2 )

countries now have ‘election law quotas’ which utilize rules governing elec-
tions to ensure some representation for women. Third, and of interest to
readers of this journal, 130 parties in 61 countries have adopted ‘political
party quotas’ in which parties set out rules or targets for the gender balance
of the individuals they place in office. Party quotas are different from consti-
tutional quotas and election law quotas in that they lack the sanctioning
power of the state. Parties simply promise to improve the gender balance
of their elected officials, but there are no official punishments for failing to
live up to these promises.
While considerable attention has been devoted to documenting the types
of quotas which have emerged around the world (e.g. Global Database,
2003; Lovenduski and Norris, 1993), to explaining why they were adopted
(e.g. Caul, 2001; Krook, 2004), and to investigating which quota provisions
are most effective (e.g. Jones, 2004), a recent symposium on the state of
gender quota research argued that the field now needs to turn to the actual
implementation of gender quotas (Squires, 2004: 55; see also Dahlerup,
2002). Work has indeed been done on the effectiveness of election law
quotas (Htun and Jones, 2002; Jones, 1998), but less systematic attention
has been paid to the factors which shape compliance with political party
quotas. Such attention seems particularly merited because, comparative
research has shown, quotas work best when there are effective sanctions for
non-compliance (Dahlerup and Friedenvall, quoted in Dahlerup and
Nordlund, 2004: 97). Even legally enforced electoral law quotas have met
with mixed success in increasing women’s representation in legislatures
(Jones, 2004). Since political party quotas lack state enforcement mechan-
isms, they seem particularly prone to failure. As two observers noted, party
rules ‘that are followed, are a huge advantage’ to women’s representation,
but rules that are not adhered to have no effect at all (Matland and
Montgomery, 2003: 32, italics in the original; see also Matland, 1998). To
date, however, we do not know what factors increase or decrease the like-
lihood that a given political party will actually implement its quota once
adopted. This article is a first step in this direction.
One reason the implementation of political party quotas is understudied
is that much research on women’s representation focuses on the national
level. Because gender quotas are relatively recent phenomena, there have
only been a limited number of national elections in which any given party
has actually implemented its quota. It is quite difficult to systematically
compare different political parties’ implementation due to the high degree
of variance both in the types of quota rules adopted among party families
and in the political contexts in which these rules are to be followed cross-
nationally.
Turning to the sub-national level allows one to avoid these problems,
however, as branches of the same political party are charged with imple-
menting identical quota rules in subunits of the same political system. As a
result, existing research on election law quota implementation has often
212
 D AV I D S O N – S C H M I C H : P O L I T I C A L P A R T Y G E N D E R Q U O TA S

focused on provincial or municipal elections (Jones, 1998, 2004; Schmidt

and Saunders, 2004). Similarly, in a federal system, studying state-level elec-
tions drastically increases the number of observations of a given party’s
quota implementation, enabling one to explain the causes of the quota’s
success or failure.
In this vein, this article examines the implementation of political party
quotas in German state legislatures during the 1990s. The German Länder
provide an excellent case in which to investigate the implementation of
party gender quotas. In the late 1980s and early 1990s four German politi-
cal parties adopted gender quotas for the country’s federal legislature, the
Bundestag, and for Germany’s 16 state legislatures, called Landtag.
Although the same parties made the same promises across states, gender
quotas had mixed success. While these quotas have on average significantly
increased women’s representation in German state legislatures, the degree
to which quotas were adhered to varied considerably across the 16 Länder.
Furthermore, Germany’s diverse states use a range of electoral systems, have
somewhat different party systems, feature varied numbers of potential
female candidates, and run the gamut from highly secular city-states to
rural, Catholic Länder. Thus the German case provides the variance needed
to explain the successful (or failed) implementation of these political party
quotas. Finally, Germany is rated very highly in the United Nation’s Gender-
Related Development Index (UNDP, 2003) and citizens there have attitudes
among the most compatible with women’s representation of any people in
the world (Inglehart and Norris, 2003). If party gender quotas are not
implemented in such a conducive setting, they seem likely to be even less
effective elsewhere.
To investigate the implementation of political party gender quotas, this
article proceeds in four steps. First, the quota specified in each party’s statute
is described. Second, using data from state legislative handbooks – in which
parliamentarians are required to publish their photograph and biographi-
cal information – the article measures to what degree women’s state-level
representation has increased in Germany since the adoption of quotas and
how representation has varied across states and parties. The third section
explains the sources of this variance, arguing that cultural characteristics of
political actors, rather than structural aspects of the political system, are
primarily responsible for the success or failure of party gender quotas. The
conclusion discusses the broader implications of these findings.

Political Party Gender Quotas in Germany

The idea of gender quotas first emerged in Germany in the 1970s when
female members of the Social Democrats (SPD) began to voice dissatis-
faction with the male domination of their party. Quotas proved contentious,
however, and they were not initially adopted. In the early 1980s the Greens
213
 PA RT Y P O L I T I C S 1 2 ( 2 )

entered the Bundestag for the first time and, in keeping with their ideo-
logical support for participatory democracy and equality, they introduced
quotas for women in their party in 1985. By 1990 the Greens promised to
give women 50 percent of all party and electoral offices. To meet their goal,
the Greens promised to use the ‘zipper system’, whereby every other slot on
their electoral list was to be filled by a woman.
This development was very well received by many of the SPD’s female
voters and the party began to fear losing them to the Greens, who were
already chipping away at the Social Democrats’ electoral base (Kolinsky,
1989: 236–7). As a result, in 1988 the SPD also adopted a gender quota,
promising gradually to improve women’s electoral representation in the
coming decade.1 Between 1988 and 1990 they promised to allot 25 percent
of their ballot positions to women; between 1990 and 1994 one-third of
slots; and after 1994 40 percent of electoral places were to go to female
candidates. The moving target was designed to allow the party time to
develop an increasingly large cadre of qualified female candidates.
After German unification in 1990, the former East German Communist
Party renamed itself the Party of Democratic Socialism (PDS) and, in an
effort to refashion itself into a new left style party, adopted the Green’s
zipper system and 50 percent quota.
The Christian Democratic Union (CDU) was slower to adopt what they
ultimately called a women’s ‘quorum’ in their party. Rank and file party
members were generally unenthusiastic about quotas, but the party leader-
ship feared losing electoral support among younger women and decided to
implement changes from above to appeal to such voters (Kolinsky, 1993:
132; Wiliarty, 2002: 174, 182). In 1985 the CDU leadership recommended
that the party send women to parliament in proportion to their numbers
within the party, about one quarter (Kolinsky, 1989: 240). Throughout the
subsequent decade, quotas were often discussed but never agreed upon
(Wiliarty, 2002). At the 1996 party congress, under pressure from CDU
Chancellor Helmut Kohl and the CDU’s Bundestag President Rita Süssmuth
(Wiliarty, 2002: 179–80), quorum requirements were finally incorporated
into the CDU’s statute, calling for the use of a two-to-one ratio when drawing
up electoral lists. The statute agreed to, however, also allowed that if nomi-
nating committees were unable to comply with this requirement they could
include fewer female candidates as long as they ‘justified’ their actions.
The Christian Democrats’ Bavarian sister party, the Christian Social
Union, has not adopted a gender quota of any kind. Similarly, the liberal
Free Democratic Party, represented in some German states, also has no
statute governing women’s representation. The CDU’s reluctance to adopt
even a limited quota and the other rightist parties’ lack of quotas are consist-
ent with the finding from across western Europe that leftist parties are more
likely to adopt gender quotas than parties of the right (Caul, 2001).
Other parties represented in some, but not all, of the Länder between
1990 and 2000 included the far-right Republicans and the German People’s
214
 D AV I D S O N – S C H M I C H : P O L I T I C A L P A R T Y G E N D E R Q U O TA S

Union (DVU), the Danish minority party, the SSW, in Schleswig-Holstein,

the STATT protest party in Hamburg, and, in Bremen, electoral lists called
the BK and the AFB. Because these parties either do not employ quotas or
are not represented outside a single state or electoral period, they are
excluded from the following analysis.

Inconsistent Implementation of Party Quotas Across the

Länder

Overall, Germany’s political party quotas have been overwhelmingly

successful in increasing women’s representation. The percentage of women
in the Bundestag rose from 17.8 percent in the late 1980s (Schindler, 1999)
to 32.8 percent in 2002 (Deutscher Bundestag, 2005). There has also been
a steady rise in the percentage of women in state legislatures. In West
Germany before unification, the average Landtag contained 16.6 percent
women (Kolinsky, 1991: 71). In 1990, the figure for united Germany was
22 percent; in 1995, women filled 30 percent of state legislative seats; and
in 2000 this rose to 31 percent.2
The Greens’ and PDS’s use of the zipper system helped them consistently
approach an equal number of men and women in their state parliamentary
delegations (see Figure 1). On average, 43 percent of Green parliamentary
party groups were women in 1990. Given that the average Green caucus
had approximately 10 members, this 43 percent figure implies that there was
only one more man than there were women in the party delegation. This
means that the zipper method functioned and the last candidate to get in
was a male. In order to take this possibility into consideration, throughout

70

60 58
% of MPs who were women

50 50 49
50 48 47
43
39 40
40
35 35
33 33

30
25 26 25
21
20
15

10

0
ta

90

95

00

00
en a

ta

PD a

SP ota

m
90

en 5
00

90

95

00
t
99

19 199
t

uo
uo

uo

uo

ru
19

19

20

20
19

19

20

19

19

20
u
1

uo
Q

Q
Q

D
en

U
Q
S

90

94

98
en

PD

PD

PD

SP SP

D
re

re

re

PD

U
C

C
19

19
re

D
G

C
D

D
SP

SP

Figure 1. Implementation of party gender quotas: Average across 16 states

215
 PA RT Y P O L I T I C S 1 2 ( 2 )

the article ‘adherence’ to a quota is defined as being within 5 percent of a

party’s goal. Using this definition, the average Green caucus met its quota in
1995 and exceeded it in 2000. On average, the former communist PDS simi-
larly met its 50 percent target in 1995 and 2000, although they too fell short
when it was introduced in 1990. The Social Democrats, on average, met
their quotas each year; in 1990 and 1995 the percentages of women repre-
sented were above agreed upon levels. The percentage of female Christian
Democratic MPs in the average party group also increased each year, even
though the CDU’s quorum was only introduced after the first two electoral
periods. However, on average, the party failed to meet its quorum in 2000.
These averages mask considerable variance, however (see Table 1). No
party met, or failed to meet, its quota in every state in which it had repre-
sentatives. For example, while the PDS filled its quota in all six states in
which it was represented in 1995, the party met its quota in only half of
these states in 1990. In 2000, the CDU managed to meet its quorum in only
3 of the 16 Länder, whereas that year the Greens filled their quota in 9 of
the 10 states in which they were represented. Just as parties inconsistently
adhered to their quotas, the percentage of women in state legislatures varied
across Germany’s states (see Figure 2). Between 1990 and 2000 the average
state legislature contained 27.6 percent women. In Baden-Württemberg, this
number fell to only 12.8 percent, whereas in Bremen the figure was three
times higher: 38.3 percent. Of course, given that different parties have
different gender quotas, much of this variance can be attributed to the
partisan composition of state legislatures.
However, when the data are disaggregated by party, an additional trend
becomes clear. In certain legislatures a ‘virtuous cycle’ appears where all
parties exceed or approach their gender quotas, whereas in other state
parliaments a ‘vicious cycle’ leaves all parties far from their goals for
women’s representation (see Table 2). In 2000 in Baden-Württemberg, for
example, the Christian Democrats, the Social Democrats and the Greens not
only all failed to meet their quotas, but these three parliamentary party
groups also contained the lowest percentage of female MPs that these three
parties had in any state. Between 1990 and 2000 in Baden-Württemberg
only one party (the Greens) actually fulfilled its quota and it did so only

Table 1. Frequency with which quotas were met

1990 1995 2000
CDU No quorum No quorum 3 of 16 states
Greens 7 of 13 states 8 of 10 states 9 of 10 states
PDS 3 of 6 states 6 of 6 states 4 of 6 states
SPD 9 of 15 states 13 of 15 states 10 of 16 states
Note: Number of states in which quota was exceeded or filled within 5 percent/number of
states in which party was represented.
Source: State legislative handbooks. Data from Rheinland-Pfalz in 1990 and 1995 are missing.

216
 D AV I D S O N – S C H M I C H : P O L I T I C A L P A R T Y G E N D E R Q U O TA S

45%

40% 37.6% 38.1% 38.3%

35.0%
35%
30.5%
30% 28.2% 28.7%
27.0% 27.6%
25.8% 25.8% 26.1%
24.7%
25% 22.2% 22.6%
19.8%
20%

15% 12.8%

10%

5%

0%
rn

en
n

nd

rg

n
lz

lt

n
en

en

n
rg

n
ha

ea

rli

ei
se

se
ur
er

le
fa
ye

bu
be

em
rla
hs
g

st
Be
tfa
nb
m
-P

es
An

ch

M
rin
Ba

am
em

ol
ac

a
m

Br
es
nd

de

H
Sa

Sa
ue

n-

-H
po
rs

H
rtt

-W
la

an
se
Th

ig
de

or
üe

in

w
ch

Br

in
-V
ie
he

es
W

he
Sa
N

rg
R
n-

hl
dr
bu

Sc
de

or
en
Ba

N
kl
ec
M

Figure 2. Average % of state legislative seats held by women 1990–2000

once; in all other instances, party quotas failed here. Brandenburg, Nieder-
sachsen and Mecklenburg-Vorpommern also had minimal success with
quotas; parties there adhered to gender quotas less than half of the time.
By contrast, in the year 2000 the state of Bremen had higher than average

Table 2. Virtuous and vicious cycles

State 1990 1995 2000 Total (%)
Baden-Württemberg 1/2 0/2 0/3 14
Brandenburg 1/3 2/2 0/3 37.5
Niedersachsen 1/2 1/2 1/3 43
Mecklenburg-Vorpommern 0/2 2/2 1/3 43
Bayern 0/2 2/2 1/2 50
Nordrhein Westfalen 1/2 1/2 2/3 57
Sachsen 1/3 2/2 2/3 62.5
Thüringen 1/3 2/2 2/3 62.5
Rheinland-Pfalz Missing Missing 2/3 67
Sachsen-Anhalt 2/3 3/3 1/3 67
Saarland 1/1 2/2 1/2 80
Hamburg 2/2 2/2 2/3 86
Schleswig-Holstein 2/2 2/2 2/3 86
Bremen 2/2 2/2 3/3 100
Hessen 2/2 2/2 3/3 100
Notes: Numbers represent the number of parties in a given state adhering to or exceeding their
gender quotas in that election divided by the number of parties having a quota in that
particular election. Adhering is defined as coming within 5 percent of the quota. Parties
include CDU, SPD and Greens (west) or PDS (east).
Source: State legislative handbooks.

217
 PA RT Y P O L I T I C S 1 2 ( 2 )

representation for women in all parties represented. The Greens and the
Social Democrats exceeded their women’s quotas and the CDU actually met
its quorum – one of only two states in which this occurred. In fact, in every
single election during the 1990s, every party with a quota in Bremen met
or exceeded it; the same was true in Hessen. Berlin, Hamburg and
Schleswig-Holstein also saw high levels of adherence to gender quotas.

Explaining Variance in the Success of Party Quotas

Why have all parties represented in some states gone over and above volun-
tary quota requirements while in other legislatures all or most parties failed
to adhere to quotas? In other words, what causes successful or failed imple-
mentation of party gender quotas? To answer these questions, it is useful to
make a distinction between the structure of a political system and agency
within it. Extensive comparative research on women and legislatures has
shown that certain political structures, first among them proportional repre-
sentation (PR) with large district and party magnitudes, are conducive to
women’s representation. However, other research has found that even
within political settings structured to women’s advantage, there need to be
individuals willing to use the system to promote women. Matland (1998),
Reynolds (1999) and Norris (2004) have all found PR electoral systems to
be more strongly associated with female representation in developed and/or
Protestant countries than elsewhere in the world where popular attitudes
are less accepting of women in political roles. The same is likely to be true
of political party gender quotas. They too are a structure conducive to
women’s representation, but one that requires actors motivated to take
advantage of the inherent possibilities for women offered – but not required
– by voluntary party quotas.
Cross-national evidence regarding the implementation of election law
quotas also highlights the importance of agency. In 1999, France passed a
‘parity’ law requiring that parties have electoral lists containing equal
numbers of men and women or be subject to a fine; at the national level,
major French parties preferred paying fines to implementing quotas (Sineau,
2002). In Latin America, Htun and Jones found election quota laws to
have only limited effects on women’s representation because many parties
implemented them in a ‘minimalist’ manner; such quotas succeeded only if
both quota-friendly political structures and ‘good-faith compliance by
political parties’ were present (2002: 32). In Costa Rica, election law quotas
only effectively increased the number of women in parliament after the
quota law was changed from requiring women to occupy at least 40 percent
of the slots on a party’s list to 40 percent of the winnable slots (Jones, 2004).
Argentina’s quota law, the Ley de Cupos, succeeded but only after women
activists from across the political spectrum took parties to court to force
them to comply with the law (Htun and Jones, 2002).
218
 D AV I D S O N – S C H M I C H : P O L I T I C A L P A R T Y G E N D E R Q U O TA S

Because there are no state sanctions for non-compliance to motivate

actors uninterested in filling party gender quotas, this type of quota is even
more dependent on quota-friendly agency within the political system for
success. The following section reviews the literature on political structures
conducive to women’s representation, depicts relevant candidate supply and
demand factors, discusses the vital importance of agency within these struc-
tures, and outlines how these variables can be measured in the German case.

Political Structures
One of the most consistent findings about women’s representation in legisla-
tive bodies is that the electoral system matters considerably (Caul, 1999;
Kenworthy and Malami, 1999). Plurality systems make it more difficult for
women to be elected as there can only be one winning candidate per district,
and while theoretically parties can divide up all their single-member district
nominations among men and women, it is very unlikely – although techni-
cally possible – that parties will choose to take seats away from popular
(usually male) incumbents. In contrast, PR – especially in the industrialized
countries – has been found to greatly increase women’s chances of election,
because it creates a positive sum game in which both male and female candi-
dates from the same party can win seats in a given electoral district
(Duverger, 1955; Henig and Henig, 2001: 95; Rule, 1987).
Germany’s federal-level electoral system is a mixed one, allotting half of
the Bundestag’s seats through a plurality system and the other half through
PR lists. Prior to the adoption of party quotas, women in the Bundestag
were four times more likely to have been elected via PR than via plurality
(Rule, 1987). Most German states have adopted this electoral system as
well. There are, however, some variations on the Länder’s electoral systems
which may influence the percentages of women elected in these states.3
The percentage of state legislature seats awarded via PR lists ranges from
100 percent in Hamburg and Bremen to only 20 percent in the Saarland.
The greater the percentage of seats awarded via PR, the greater the likeli-
hood that parties will be able to meet their quotas. Prior research on PR
has also found that the larger the district magnitude, the more likely the
results will be proportional to a given group’s percentage in the population
(Engstrom, 1987; Henig and Henig, 2001: 95–6; Rule, 1987). In Germany,
the number of seats on lists ranges from 121 in Hamburg to only 10 in the
Saarland. The longer the list, then, the more favourable the ratio of female
candidates is hypothesized to be. Conversely, in countries where plurality
electoral systems are used, women have been found to be more successful
in districts with smaller populations (Rule, 1990). In Germany, the number
of constituents represented by a direct mandate ranges from 142,000 in
Nordrhein Westfalen to 26,000 in the Saarland. The smaller the plurality
district size, then, the more successful parties should be in meeting their
quotas.
219
 PA RT Y P O L I T I C S 1 2 ( 2 )

Where PR is combined with preferential lists, voters can determine the

order of candidates’ names on the party list. Findings on gender and pref-
erential lists have been varied. Some scholars have found evidence that this
electoral system may work against women because voters choose to bump
new female candidates down the list in favour of established ‘old boys’
(Ellis, 2002; Jones, 1998). However, others have documented incidents of
feminist voters using preferential lists to place women at the top of the list
(Henig and Henig, 2001: 99; Norris, 1985). Only one German state –
Bayern – uses such preferential lists and its experience is similarly mixed.
One study investigated the fate of female candidates in Bavarian state elec-
tions and found that half of them were moved down the list, one-third were
bumped up and the rest were unchanged (Hoecker, 1994: 73). A dummy
variable is used in the analysis below to assess the effect of Bayern’s pref-
erential list electoral system.
In addition to the electoral system, the party system may also exert an
influence on the success of gender quotas. Where party competition is high,
parties may fight their battles on key issues of the day, ignoring quotas and
running male candidates who are perceived to be safer bets. Where one
party or coalition is clearly dominant, however, there may be more leeway
for parties to pay attention to gender quotas. Opposition parties may try to
woo female voters away from the dominant party by stressing their commit-
ment to gender quotas. Such strategic action, often on the part of a smaller
party facing a large rival or a party having suffered a large defeat, accounts
for the initial adoption of many gender quotas (Caul, 2001; Krook, 2004:
60; Lovenduski and Norris, 1993: 9; Matland and Studlar, 1996). In the
following analysis, party competition is measured in the west by taking the
average difference in the vote share received by the Christian Democrats
and liberals, and the vote share received by the Social Democrats and the
Greens, in elections across the 1990s. In the three-party eastern system, the
vote share of the smallest of the three parties represented (CDU, PDS, SPD)
was subtracted from the vote share of the largest of the three parties.
A party’s degree of centralization and the institutionalization of their
nomination process also influences women’s chances of being elected (Caul,
1999; Henig and Henig, 2001: 49–50; Lovenduski and Norris, 1993: 13–14,
327). In highly centralized parties, leaders can, if they choose, easily decide
to place more women on the ballot. Additionally, the more institutionalized
the nomination process is, the easier it is for outsiders, such as women, to
determine what has to be done to be nominated for a slot of the ballot.
German laws governing candidate selection are identical across the country.
The law requires that candidates for federal and state elections be chosen
in a ‘democratic’ manner by party members – an assembly of members in
the electoral district for plurality seats and at the state or regional level for
PR lists. Correspondence with the various parties’ women’s auxiliary groups
across the Länder suggests that the de facto process of candidate selection
within each party is relatively constant from state to state. As a result, this
220
 D AV I D S O N – S C H M I C H : P O L I T I C A L P A R T Y G E N D E R Q U O TA S

variable does not seem capable of accounting for varying quota adherence
across Germany for most of the 1990s and will not be considered in the
analysis below.
However, this variable is relevant to eastern German Länder in 1990.
There, state elections were held immediately following unification, while
political parties were still being formed. Interviews with state legislators
nominated and subsequently elected in this period paint the picture of an
extremely chaotic nomination process (Davidson-Schmich, 2005). Party
meetings were held on an evening prior to the election, individuals declared
their candidacy and the people who gave the most convincing speeches were
selected for the ballot. Scant attention was paid to gender (or any other
strategic characteristics) when lists were drawn up. Because potential female
candidates may have had childcare or other family responsibilities on the
evenings the lists were drawn up, or because women’s voices may not have
been as persuasive given gender stereotypes, this decentralized, uninstitu-
tionalized nomination process was not conducive to women’s representa-
tion. A dummy east/west variable is used to capture this variation on
recruitment in the analysis below.
Even with a highly institutionalized nomination process, however, if there
are few women who are available to select as candidates, the party most
committed to women’s representation will still have difficulty fulfilling its
quota. Indeed, this problem finding a supply of candidates plagued the
Greens in the early years of their women’s quota (Kolinsky, 1991; Norris
and Lovenduski, 1993: 405). Candidates in Germany are usually drawn
from four main areas: internal party offices, local elected offices, related
interest groups and the public sector (Kaltefleiter, 1976; Patzelt, 1995;
Saalfeld, 1997; von Beyme, 1986). In order to document potential variations
in the supply of female candidates across the German Länder, each of these
recruitment channels is investigated below.
The parties in Germany recognize the importance of developing a cadre
of women party officials for reaching their gender quotas. The Greens, the
Social Democrats and the Christian Democrats have all developed quotas
not only for their electoral lists but also for internal party offices. While
national party initiatives to promote women in internal offices apply across
the Länder, they may meet with varied success from state to state. To deter-
mine whether this has been the case or not, the percentages of women in
top party offices were measured. These offices included the state head and
deputy head of the party, its general secretary, treasurer and members of its
board of directors. The higher the percentage of women in internal party
offices, the more likely there will be potential female candidates for a state
party to draw on and, in turn, the more successful gender quotas are
expected to be.
Similarly, the higher the number of women serving in local councils, in
public administration4 and in key interest groups in a given state, the higher
the number of qualified candidates there will be for state-level elections and
221
 PA RT Y P O L I T I C S 1 2 ( 2 )

the more likely quotas are to succeed. To measure the number of women
serving in local elected office, the average number of women serving on
city/county councils in the two largest cities and two largest counties in a
given state was taken. The number of women in public service was calcu-
lated as a percentage of state-level employees who were women; these data
came from state statistical offices. To gauge the number of women in interest
group leadership positions, the gender of the head and deputy head of three
major interest groups in each state was determined. These groups included
the teacher’s union (Gewerkschaft Erziehung und Wissenschaft), the police
union (Gewerkschaft der Polizei) and an environmental group (BUND);
each has close ties to one of the major political parties, i.e. the Social
Democrats, the Christian Democrats and the Greens/PDS, respectively.
A large supply of candidates would be even more conducive to quota
success if there were a high demand for candidates. In settings where there
is a high level of legislative turnover, many new opportunities are created
for women and other novice candidates (Norris, 2004: 189; Norris and
Lovenduski, 1995: 29; Welch and Studlar, 1990: 401). In contrast, where
turnover is low, established (usually male) incumbents may leave little room
for newcomers. It will be harder for proponents of women’s quotas to fulfil
quotas by displacing incumbents than it will be for pro-women agents in
Länder where there are many vacancies in the state legislature. Thus, the
percentage of legislative seats that turn over from year to year may help
explain quota adherence across the German states. In order to determine
the amount of turnover that German state legislatures have seen in the past
decade, the tenure of each MP in 2000 was measured and averaged for each
Land. The shorter the tenure of the average deputy, the more turnover there
was in that legislature in the past decade and the more likely it would seem
that parties would be able to meet their quotas by introducing new, female
candidates.

Agency within Political Structures

While the above electoral and party systems – as well as the supply and
demand factors considered – are certainly not unimportant for the success
or failure of political party gender quotas, it is also vital to consider agency
within this structure. As such quotas are voluntary, if there is no one actively
recruiting women or if the people who select among potential candidates
are hostile to gender quotas, female candidates may be overlooked or
excluded – even in Länder with conducive electoral and party systems, a
large supply of potential candidates and a demand for new candidates.
Often those most active in promoting women’s candidacy are other
women. As Lovenduski and Norris observed in their work on women and
political parties in western Europe, ‘there [was] no party in which efforts
to nominate more women have occurred without an intervention by women
making claims’ (1993: 14). Indeed, the number of women in a given party’s
222
 D AV I D S O N – S C H M I C H : P O L I T I C A L P A R T Y G E N D E R Q U O TA S

leadership is a strong predictor of whether a given party will adopt gender

quotas in the first place (Caul, 2001). Thus, in Germany, one set of import-
ant agents who are likely to recruit women and promote gender quotas are
women in the state executive. Female Prime Ministers (Ministerpräsidentin)
and Cabinet Ministers are significant figures within their parties and can
play an important role in identifying, mentoring and promoting other
women for vacant legislative seats. In order to assess the impact of these
important promoters of female candidates on the implementation of gender
quotas, the total number of Cabinet positions in a given state over the 1990s
was counted, as was the number of women who occupied these seats during
the decade. This percentage is used in the analysis below.
Women’s organizations outside the executive or legislative branch have
also been found to play a role in the establishment of gender quotas (Antić
and Gortnar, 2004: 77–8; Jones, 2004; Krook, 2004: 63). In a study of
established democracies, Caul (1999) found that the presence of female
activists within the rank-and-file of a given party helped ensure women
would be promoted. Women’s organizations have therefore been hypothe-
sized to play a role in pressing parties to adhere to their gender quotas
(Dahlerup and Nordlund, 2004: 97; for an account of such an incident in
Germany, see Kolinsky, 1993: 138, 140). One possible explanation for
varying quota implementation in Germany, then, may be the presence or
absence of active women’s groups across the Länder. The women’s groups
most concerned with quota implementation are the women’s auxiliary
organizations in the large political parties: the Christian Democratic Frauen
Union, the Social Democratic Arbeitsgemeinschaft sozialdemokratische
Frauen (AsF) and the Party of Democratic Socialism’s LISA. The head of
each state branch of these organizations was contacted in order to deter-
mine what actions these groups are undertaking to promote female candi-
dates in each Land. The responses revealed that there is little state-to-state
variance among the types of activities that women’s auxiliary organizations
undertake to promote female candidates; as a result, this variable is
excluded in the analysis below.
While high-ranking women or women’s organizations may advocate
quota fulfilment, it is also necessary that ‘gatekeepers’, or those whose job
it is ultimately to select among competing candidates, are sympathetic to
their demands. This is because gatekeepers have the ultimate responsibility
for creating party lists and choosing direct mandate candidates in accord-
ance with their party’s statute (Norris, 1997; Norris and Lovenduski, 1995;
see also Dahlerup, 2002: 3; Matland and Montgomery, 2003: 21). As
mentioned above, in Germany these gatekeepers are assemblies of rank-and-
file party members who must ultimately approve single-member district
nominees and party lists. Since there is a very strong overall correlation
between popular attitudes towards gender equality and women’s political
representation (Inglehart and Norris, 2003), it is likely that such attitudes
will also influence gatekeepers’ willingness to press for adherence to party
223
 PA RT Y P O L I T I C S 1 2 ( 2 )

gender quotas. If there is a broad sense within a state among men and
women that party quotas are worthwhile and should be adhered to if at all
possible, such quotas are likely to succeed. If in other states, in contrast,
both female and male gatekeepers have little interest in adhering to quotas,
they are likely to fail.
As countries become more economically developed, citizens have been
found to become more supportive of women serving in political office
(Inglehart and Norris, 2003). Clearly, the level of economic development
does not vary much across Germany, but the country does contain large
urban centres such as Berlin and Frankfurt, as well as more traditional rural
areas such as the Black Forest. Studies of women in German local politics
have found fewer women represented in smaller, rural communities than in
large cities (Duverger, 1955: 77–8; Hoecker, 1994: 67; Kolinsky, 1989:
223). It therefore seems likely that gatekeepers sympathetic to quota adher-
ence are more likely to be found in Germany’s more urbanized states than
they are in large, thinly populated Länder. Data from Germany’s Federal
Statistical Office, which measures the percentage of a given state’s GDP
derived from agriculture, are used to test this hypothesis.
Even among countries with the same level of development, however,
support for women’s political participation remains varied, with attitudes
diverging along religious lines. Numerous studies from western Europe and
beyond have found Catholicism to be associated with lower levels of
women’s legislative representation than Protestantism (Norris, 1997:
217–18; 2004: 207; Paxton, 1997; Rule, 1987; 1994: 20). Prior to the intro-
duction of quotas, more women were represented in northern (Protestant)
German state parliaments than in the Catholic, southern legislatures
(Kolinsky, 1993: 137). In Germany’s strongly Catholic states, then, reaching
women’s quotas is unlikely to be seen as a priority by either male or female
gatekeepers. Their counterparts in heavily Protestant states may be more
proactive in working to fill voluntary quotas. To assess Catholicism, the
German Bishops’ Conference’s measure of the percentage of state popu-
lations that are Catholic was utilized. In sum, in rural Catholic states with
few high-ranking women in state politics, implementing party gender quotas
is likely to be given low priority by party gatekeepers – even if state-level
political structures may be conducive to this end.

Results

This section first discusses the bivariate relationships between the above-
mentioned variables and the success of gender quotas in the German Länder
in the 1990s and then turns to multivariate analysis. While the bivariate
relationships confirm the importance of both structure and agency to the
success of voluntary gender quotas, the multivariate analysis underscores
the importance of agency.
224
 D AV I D S O N – S C H M I C H : P O L I T I C A L P A R T Y G E N D E R Q U O TA S

Bivariate Results
The German political system plays an important part in the ability of parties
to fulfil their gender quotas; strong correlations between state electoral
systems and the success of quotas were observed (see Table 3). The larger
the percentage of seats filled by PR, the more successful the gender quotas
were. Similarly, the longer electoral lists were in terms of the number of
candidates, the more likely parties were to meet their quotas. Finally, the
dummy variable for Bayern produced a negative correlation, suggesting that
voters may move women’s names down electoral lists in favour of male
candidates. This finding is consistent with an earlier study (Hoecker, 1994:
73). In the 14 states which had candidates elected via direct mandates as
well as PR lists, large districts indeed proved detrimental to female candi-
dates. The German case also provides evidence that competitive electoral
politics hindered the success of gender quotas. The more competitive state
party systems were, the less successful quotas were.
As suggested by the broader comparative politics literature, decentralized,
uninstitutionalized processes of candidate recruitment made it difficult for
eastern German parties to meet their gender quotas in the wake of German
unification. There was a negative relationship between eastern Germany’s
initial decentralized nomination process and quota fulfilment there.
The results in terms of the supply of candidates did not always conform
to the expectations raised by the broader literature, however. While male
candidates in Germany have traditionally come from four different recruit-
ment channels – inner-party office, local political office, the public service
and related interest groups – the same does not seem to be the case for

Table 3. Bivariate correlations

Correlation with state-level
Variable success of quotas
PR electoral system 0.36
Length of party list 0.20
Preferential list (Bayern) –0.17
Number of citizens per direct mandate –0.39
Competitiveness of party system –0.24
Non-institutionalized candidate selection –0.24
Women in inner-party office 0.20
Women in local political office 0.19
Women in interest groups 0.02
Women in public service –0.31
Legislative turnover –0.11
Women in state-level executive office 0.34
Catholic –0.11
Rural –0.72
Note: See text for sources.

225
 PA RT Y P O L I T I C S 1 2 ( 2 )

women. There was a positive correlation between the number of women in

key inner-party offices and the success of quotas. The same was true for
local political office. The correlation between the number of women in
leading interest group positions in a given state and the success of gender
quotas in that state was negligible, however. Most surprising, the relation-
ship between the number of women in public service in a given state and
the functioning of quotas there was actually negative. It appears, then, that
German political parties are drawing on women within their own ranks,
either those serving in inner-party office or those who have had local politi-
cal careers, rather than seeking female candidates elsewhere. These findings
are consistent with Welch’s (1978) evidence that there are fewer women in
state legislatures than the eligibility pool would permit.
The results in terms of turnover were also surprising. There was a
negative relationship between legislative turnover and the success of gender
quotas. At first glance, this result appears counter-intuitive. As older, usually
male, veterans leave state political careers, they could easily be replaced by
female candidates, making it feasible for parties to meet their gender quotas.
That they did not do this in Germany may stem from some case-specific
features. The 1990s brought about a change in the make-up of the national-
level governing coalition; when the Red–Green Schroeder government took
office in Berlin it was able to provide federal- and European-level positions
for loyal Social Democrats and Greens. In an effort to send a signal about
the parties’ commitment to women, many of those who were tapped for
these offices were women holding state-level political office. As a result, the
pool of female candidates in strongly Social Democratic or Green Länder
may have been depleted, causing open seats to be filled by male candidates.
As expected, in addition to these structural aspects of the electoral and
party systems, there was a strong bivariate relationship between the success
of voluntary gender quotas and certain agents within the German Länder.
Women in executive positions were positively correlated with the success of
party gender quotas in the 1990s. Furthermore, as predicted, the more
Catholic a given state, the less successful quotas were. The strongest bivari-
ate relationship of all was the positive correlation (0.72) between agricul-
tural economies and the failure of women’s quotas.

Multivariate Analysis
Multivariate analysis – an Ordinary Least Squares regression using the
dependent variable described in Table 2 – underscores the importance of
these latter variables (see Table 4). In the model used, a state’s electoral
system was measured using the percentage of seats allotted via PR lists
rather than the length of the list or the size of single-member districts; all
of the other variables mentioned above were included in the equation.
The only structural variable which proved significant was the dummy
variable used to capture Bayern’s preferential list system; interestingly,
226
 D AV I D S O N – S C H M I C H : P O L I T I C A L P A R T Y G E N D E R Q U O TA S

however, this was significant in a positive direction. This either indicates

that preferential lists are being used by feminist voters to privilege female
candidates or this variable is capturing some other aspect of Bavarian
politics.5
The electoral system and the competitiveness of the party system were
not significant in multivariate analysis. Similarly, the variable designed to
capture the decentralized nature of the candidate recruitment process in
eastern Germany following the fall of the Berlin Wall did not prove to be
significant. Nor were most of the variables dealing with the supply of female
candidates. The percentages of women in local government, interest groups
and party offices had no effect on quota implementation. The percentage of
women in public service was weakly significant (t.10) in a negative direction.
Although German male candidates are often recruited from the public
service, a strong female presence in state-level public sectors does not help
Table 4. OLS results
Coefficient (standard error)
Intercept** 2.553
(1.26)
Electoral system –0.002
(0.003)
Party competition –0.004
(0.008)
Bayern** 1.166
(0.581)
East –0.171
(0.170)
Inner-party office 0.673
(1.078)
Local office –0.017
(0.014)
Interest group –0.088
(0.104)
Public service* –0.011
(0.007)
Turnover –0.015
(0.105)
Women in executive –0.002
(0.008)
Catholic** –0.012
(0.006)
Agrarian*** –0.313
(0.098)
Notes: Adjusted R2: 0.557.
*** 95 percent confidence interval.
** 90 percent confidence interval.
* 80 percent confidence interval.

227
 PA RT Y P O L I T I C S 1 2 ( 2 )

parties fill their quotas. Finally, the degree of turnover in a given legislature
did not prove significant in the regression results.
As expected, the success of voluntary gender quotas in the German states
hinged not on the political structure of these Länder, but rather on the
willingness of individuals within the system to act on the opportunities
inherent in these structures. Most significant in the multivariate analysis
were the variables measuring both Catholicism and agricultural economies.
The presence of women in state-level executives did not significantly
increase quota adherence, however.

Discussion

This article has shown that, in the German case, political party quotas did
make a difference, despite their voluntary nature. Women’s representation
at the federal and state level in Germany doubled over the course of the first
decade that quotas were in effect. Furthermore, the findings presented here
suggest that just as party ideology shapes the willingness to adopt gender
quotas, it also makes a difference in terms of whether parties actually imple-
ment their quotas. Left-wing parties, especially new-left parties, set the
highest quotas and came much closer to meeting them than did the right-
wing Christian Democrats, whose women’s quorum was more modest.
However, this article also reveals that voluntary quotas are not always
adopted with equal enthusiasm across various branches of the same parties.
In some states, all parties met or exceeded their quotas, while in others, all
parties failed to meet their party family’s quota. Certain structural factors
such as PR with large district magnitude, plurality with small electoral
districts, institutionalized candidate recruitment processes and large
numbers of women in party and other elected offices, were indeed corre-
lated to the success of voluntary women’s quotas – just as they increase
women’s representation in settings without quotas. In the final analysis,
however, political structures conducive to women’s representation were no
guarantee that quotas would be implemented. Instead, the attributes of the
gatekeepers who select among promising candidates were the key factor in
determining whether or not voluntary gender quotas actually succeeded.
Catholicism and a more agricultural economy created an environment
inhospitable to gender quotas in some German states. These findings are
consistent with an established literature that finds PR electoral systems are
used to increase women’s representation more in developed Protestant
countries than they are in less economically developed Catholic areas. This
evidence is particularly striking considering the case under examination
here; compared to other countries in the world, rural Catholic states in
Germany are fairly industrialized and secular. If these variables inhibited the
successful implementation of party gender quotas in the Länder examined
here, they are likely to further reduce the effectiveness of party gender
228
 D AV I D S O N – S C H M I C H : P O L I T I C A L P A R T Y G E N D E R Q U O TA S

quotas in other settings. Future research on the implementation of party

gender quotas at the national and sub-national level in countries across the
globe can confirm whether this is indeed the case.
A final avenue for future research involves the mechanisms through which
Catholicism inhibits, and urbanism promotes, gender quota implementa-
tion. In rural or Catholic areas, do women internalize taboos against female
participation in politics, never considering a run for office? Or do parties in
such states have an infrastructure which hinders women’s political partici-
pation through more mundane means, such as holding meetings during
children’s bedtimes? Or do women attempt to run, only to be thwarted by
biases on the part of gatekeepers? Or are other mechanisms at work?
Answering these questions will help advocates of party gender quotas
improve their implementation in such areas.

Notes

Earlier versions of this article were presented at annual meetings of the Southern
Political Science Association and the German Studies Association. Thanks to
audience members and to anonymous reviewers for their constructive comments.
Thanks also to Markus Thiel, Martin Kaiser, Jason Niggley and Daniel Pedreira for
their research assistance and to the University of Miami for its financial support.

1 This pattern of the Green party serving as a quota ‘contigation’ was replicated
elsewhere in Western Europe during this period (Matland and Studlar, 1996).
2 In Germany’s federal system, state-level elections are held at varying intervals
across states due to different terms of office and occasional government collapses.
Thus, there is no one year that is ‘election year’ in all states. In the following
section, when I refer to the ‘1990’, ‘1995’ and ‘2000’ elections I am referring to
elections for the legislative periods that encompass each of these years. For
example, in Mecklenburg-Vorpommern the ‘1995’ election was actually held in
1994 for a term running 1994–8; and the ‘2000’ election was held in 1998. All
data come from the aforementioned legislative handbooks.
3 All electoral system data were gathered from the states’ Electoral Offices
(Landeswahlämter).
4 German law stipulates that all public-sector employees are allowed to take time
off from their job to serve in elected office. Moreover, should they be voted out
of office at some point in the future, they are guaranteed the right to return to
their old job (or to an equivalent position).
5 Just as Texas is in the US, Bayern is widely considered an outlier among the
German states in terms of its character.

References

Antić, Milica G. and Maruša Gortnar (2004) ‘Gender Quotas in Slovenia: A Short
Analysis of Failures and Hopes’, European Political Science 3: 73–9.

229
 PA RT Y P O L I T I C S 1 2 ( 2 )

Caul, Miki (1999) ‘Women’s Representation in Parliament: The Role of Political

Parties’, Party Politics 5: 79–88.
Caul, Miki (2001) ‘Political Parties and the Cross-National Adoption of Candidate
Gender Quotas: A Cross-National Analysis’, Journal of Politics 63: 1214–29.
Dahlerup, Drude (2002) ‘Quotas – A Jump to Equality? The Need for International
Comparisons of the Use of Electoral Quotas to Obtain Equal Political Citizenship
for Women’. Paper prepared for workshop hosted by the International Institute
for Democracy and Electoral Assistance (IDEA), 25 September, Jakarta, Indonesia.
Dahlerup, Drude and Anja Taarup Nordlund (2004) ‘Gender Quotas: A Key to
Equality? A Case Study of Iraq and Afghanistan’, European Political Science 3: 91–8.
Davidson-Schmich, Louise K. (2005) Becoming Party Politicians: Eastern German
State Legislators in the Decade since Democratization. Notre Dame, IN: University
of Notre Dame Press.
Deutscher Bundestag (2005) ‘Zusammensetzung des Deutschen Bundestages’.
Available at: http://www.bundestag.de/mdb15/mdbinfo/zusammen.html (accessed
18 February 2005).
Duverger, Maurice (1955) The Political Role of Women. Paris: UNESCO.
Ellis, Melody (2002) ‘Gender Disparity in Representation: The Effect of Preference
Voting in Italy and Japan’. Paper presented at the Annual Meeting of the American
Political Science Association.
Engstrom, Richard (1987) ‘District Magnitude and the Election of Women to the
Irish Dail’, Electoral Studies 6: 123–32.
Global Database of Quotas for Women (2003) Available at: www.quotaproject.org.
Henig, Ruth and Simon Henig (2001) Women and Political Power. New York:
Routledge.
Hoecker, Beate (1994) ‘The German Electoral System: A Barrier to Women?’, in
Wilma Rule and Joseph F. Zimmerman (eds) Electoral Systems in Comparative
Perspective: Their Impact on Women and Minorities, pp. 65–78. Westport, CT:
Greenwood Press.
Htun, Mala N. and Mark P. Jones (2002) ‘Engendering the Right to Participate in
Decision-Making: Electoral Quotas and Women’s Leadership in Latin America’,
in Nikki Craske and Maxine Molyneux (eds) Gender and the Politics of Rights
and Democracy in Latin America, pp. 32–56. New York: Palgrave.
Inglehart, Ronald and Pippa Norris (2003) Rising Tide: Gender Equality and
Cultural Change Around the World. Cambridge: Cambridge University Press.
Jones, Mark P. (1998) ‘Gender Quotas, Electoral Laws, and the Election of Women’,
Comparative Political Studies 31: 3–21.
Jones, Mark P. (2004) ‘Quota Legislation and the Election of Women: Learning from
the Costa Rican Experience’, Journal of Politics 66: 1203–23.
Kaltefleiter, Werner (1976) ‘The Recruitment Market of the German Elite’, in Heinz
Eulau and Moshe M. Czudnowski (eds) Elite Recruitment in Democratic Nations,
pp. 239–62. New York: Sage.
Kenworthy, Lane and Melissa Malami (1999) ‘Gender Inequality in Political Repre-
sentation: A Worldwide Comparative Analysis’, Social Forces 78: 235–68.
Kolinsky, Eva (1989) Women in West Germany: Life, Work, and Politics. New York:
Berg.
Kolinsky, Eva (1991) ‘Political Participation and Parliamentary Careers: Women’s
Quotas in West Germany’, West European Politics 14: 56–72.

230
 D AV I D S O N – S C H M I C H : P O L I T I C A L P A R T Y G E N D E R Q U O TA S

Kolinsky, Eva (1993) ‘Party Change and Women’s Representation in Unified

Germany’, in Joni Lovenduski and Pippa Norris (eds) Gender and Party Politics,
pp. 113–45. Thousand Oaks, CA: Sage.
Krook, Mona Lena (2004) ‘Gender Quotas as a Global Phenomenon: Actors and
Strategies in Quota Adoptions’, European Political Science 3: 59–65.
Lovenduski, Joni and Pippa Norris (1993) Gender and Party Politics. Thousand
Oaks, CA: Sage.
Matland, Richard E. (1998) ‘Women’s Representation in National Legislatures:
Developed and Developing Countries’, Legislative Studies Quarterly 23: 109–25.
Matland, Richard E. and Kathleen A. Montgomery (2003) Women’s Access to
Political Power in Post-Communist Europe. Oxford: Oxford University Press.
Matland, Richard E. and Donley T. Studlar (1996) ‘The Contagion of Women
Candidates in Single-Member District and Proportional Representation Electoral
Systems: Canada and Norway’, Journal of Politics 58: 707–33.
Norris, Pippa (1985) ‘Women’s Legislative Participation in Western Europe’,
Western European Politics 8: 90–101.
Norris, Pippa (1995) Political Recruitment: Gender, Race, and Class in the British
Parliament. Cambridge: Cambridge University Press.
Norris, Pippa (1997) Passages to Power: Legislative Recruitment in Advanced
Democracies. Cambridge: Cambridge University Press.
Norris, Pippa (2004) Electoral Engineering: Voting Rules and Political Behavior.
Cambridge: Cambridge University Press.
Norris, Pippa and Joni Lovenduski (1993) ‘“If Only More Candidates Came
Forward”: Supply-Side Explanations of Candidate Selection in Britain’, British
Journal of Political Science 23: 373–408.
Patzelt, Werner J. (1995) Abgeordnete und Ihr Beruf. Berlin: Akademie Verlag.
Paxton, Pamela (1997) ‘Women in National Legislatures: A Cross-National
Analysis’, Social Science Research 26: 442–64.
Reynolds, Andrew (1999) ‘Women in the Legislatures and Executives of the World:
Knocking at the Highest Glass Ceiling’, World Politics 51: 547–72.
Rule, Wilma (1987) ‘Electoral Systems, Contextual Factors and Women’s Oppor-
tunity for Election to Parliament in Twenty-three Democracies’, Western Political
Quarterly 40: 477–98.
Rule, Wilma (1990) ‘Why More Women are State Legislators: A Research Note’,
Western Political Quarterly 43: 437–48.
Rule, Wilma (1994) ‘Women’s Underrepresentation and Electoral Systems’, PS:
Political Science and Politics 27: 689–92.
Saalfeld, Thomas (1997) ‘Professionalization of Parliamentary Roles in Germany:
An Aggregate Level Analysis 1949–94’, Journal of Legislative Studies 3: 32–54.
Schindler, Peter (1999) Datenhandbuch zur Geschichte des Deutschen Bundestages
1949–1999: Band I. Baden Baden: Nomos.
Schmidt, Gregory D. and Kyle L. Saunders (2004) ‘Effective Quotas, Relative Party
Magnitude, and the Success of Female Candidates: Peruvian Municipal Elections
in Comparative Perspective’, Comparative Political Studies 37: 704–24.
Sineau, Mariette (2002) ‘Institutionalizing Parity: The French Experience’, Inter-
national Institute for Democracy and Electoral Assistance [www.idea.int].
Squires, Judith (2004) ‘Gender Quotas: Comparative and Contextual Analyses’,
European Political Science 3: 51–8.

231
 PA RT Y P O L I T I C S 1 2 ( 2 )

United Nations Development Programme (2003) Human Development Report.

Available at: http://hdr.undp.org/reports/global/2003/.
von Beyme, Klaus (1986) ‘The Role of Deputies in West Germany’, in Ezra N.
Suleiman (ed.) Parliaments and Parliamentarians in Democratic Politics,
pp. 154–75. New York: Holmes and Meier.
Welch, Susan (1978) ‘Recruitment of Women to Public Office: A Discriminant
Analysis’, Western Political Quarterly 31: 372–80.
Welch, Susan and Donley T. Studlar (1990) ‘Multi-Member Districts and Represen-
tation of Women: Evidence from Britain and the United States’, Journal of Politics
52: 391–412.
Wiliarty, Sarah Elise (2002) ‘Bringing Women into the Party: The Christian Demo-
cratic Union as a Corporatist Catch-All Party’. Dissertation. University of Cali-
fornia, Berkeley. UMI Dissertation Number 3082460.

LOUISE K. DAVIDSON-SCHMICH is an Assistant Professor of Political Science at

the University of Miami in Coral Gables, Florida, USA. She is the author of the forth-
coming book Becoming Party Politicians: Eastern German State Legislators in the
Decade Following Democratization (University of Notre Dame Press).
ADDRESS: University of Miami, Box 248047, Coral Gables, FL 33124–6534, USA.
[email: [email protected]]

Paper submitted 13 September 2004; accepted for publication 4 March 2005.

232